Monetary policy into the new millennium
Too often, both in the Reserve Bank and, I suspect, in financial markets, we are inclined to lose sight of the longer-term, or strategic, issues of relevance to monetary policy. We get preoccupied with whether the Official Cash Rate should really have been set at 4.50 per cent rather than at 4.25 per cent, as if this detail had any enduring significance for the real economy and the lives of ordinary New Zealanders. I don't want to imply that the short-term tactics are irrelevant - far from it - but they are very much less relevant than the longer-term issues, particularly when the Official Cash Rate itself is explicitly up for review on a regular basis.
So today I want to address the question, What can New Zealanders expect of monetary policy in the long term?
The monetary policy framework is profoundly democratic
The first point to note in answering that question is that the answer depends on what Parliament wants monetary policy to deliver. It is sometimes suggested that the Reserve Bank seems to be a law unto itself, and that Don Brash seems to run the country. Nothing could be further from the truth. In New Zealand, the Reserve Bank is subject to an Act of Parliament, and Parliament can change that legislation by simple majority. This is also true in Australia, the United Kingdom, the United States and many other countries, of course, but it is not now true in much of Europe (where the European Central Bank is a creature of the Maastricht Treaty) or in those countries where the central bank has been given essentially unlimited independence under the constitution.
Moreover, that Act of Parliament requires that, within the general directive prescribed by the Act, the Governor of the Bank must sign a written agreement with the Treasurer setting out quite explicitly what the Governor must achieve during his or her term of office, the so-called Policy Targets Agreement. That Agreement can be changed by mutual agreement at any time during the Governor's term of office, and can be over-ridden by Order-in-Council if the Government wants a change and the Governor does not.
So what monetary policy will deliver in the years ahead is essentially a matter for the public, through Parliament, to determine. The Reserve Bank and I as the Bank's Governor are held accountable for our performance in delivering the objectives which Government and Parliament have chosen, through a requirement to produce regular Monetary Policy Statements and to appear regularly before the Finance and Expenditure Committee of Parliament.
Having said that, as long as the present Act remains in place at least, the Government can not change the objective laid down for monetary policy without telling the public. In other words, any change in the objective must be totally transparent and, given that inflation is essentially a form of theft, that seems entirely appropriate. All this seems profoundly democratic, and makes it clear that the Reserve Bank and Don Brash are servants of Parliament.
Since 1989, the Reserve Bank Act has made it clear that monetary policy must be used to achieve and maintain `stability in the general level of prices', with the meaning of this defined in the Policy Targets Agreement. The present Agreement, signed when I was re-appointed Governor at the end of 1997, defines `stability in the general level of prices' as meaning that monetary policy should be aimed at keeping inflation as measured by the CPI excluding credit charges, or CPIX, between 0 and 3 per cent over each 12-monthly period, though the Agreement also recognises that there will be some circumstances where inflation may legitimately be allowed to fall outside that range.
But what, I am often asked, happens after the election? Fundamentally, that depends on what the next Parliament decides. It is not inevitable that anything (related to monetary policy) will change after the election. Almost all political parties are on record as supporting the present legislation, and indeed supporting the present 0 to 3 per cent target. While the Policy Targets Agreement could be changed by agreement between Governor and any new Treasurer, there should be no presumption that that will occur. Consistency and continuity are attractive in monetary policy as elsewhere, and I shall certainly not be promoting a change. It is perhaps worth noting that other central banks with a price stability objective and a quantitative inflation target all have targets that fall within the 0 to 3 per cent range - and to the best of my knowledge, there are none that fall outside it. But, to repeat, it is for Parliament to decide in the final analysis.
The present Policy Targets Agreement
What does the present Policy Targets Agreement imply about the way in which the Reserve Bank will formulate monetary policy in the years ahead? Without going into great detail, it means that we will be aiming to keep CPIX inflation close to the mid-point of the 0 to 3 per cent target one to two years ahead.
We aim close to the mid-point because we recognise that, in an uncertain and fast-changing world, aiming at the mid-point maximises the chance that we will be able to keep inflation in the target range. Aiming near either edge of the target would run too high a risk that unexpected shocks would throw us outside the target.
And we aim to keep inflation near the mid-point not this quarter or next, but in one to two years' time, because we recognise that monetary policy has most of its impact on trend inflation over this kind of time horizon.
It follows that in implementing policy we would in principle be willing to accept the prospect of inflation outcomes close to the edges of the target, and conceivably even outside the target, in the next few quarters if we were reasonably confident that inflation would be back near the centre of the target within one to two years. Clearly, we will not lightly sanction breaches of the agreed target - indeed, we will do our utmost to avoid such breaches, of either side of the target - but to pretend that we can at all times and in all circumstances keep inflation within the target would be to promise more than we can deliver. Indeed, we learnt that in the mid-nineties.
As an aside, we are reasonably certain that headline inflation, the CPI including interest rates, will fall below zero for several quarters this year. In the 12 months to December 1998, headline inflation was only 0.4 per cent, the lowest level for several decades. We expect the 12 months `increase' to be negative very shortly.
But it is important that nobody misunderstands the significance of this. Having the headline inflation figure go below zero is not a breach of the Policy Targets Agreement, any more than it was a breach when the headline figure rose to 4.6 per cent in mid-1995. The inflation target in the Policy Targets Agreement is explicitly defined in terms of the CPI excluding interest rates, for the very good reason that an interest-inclusive CPI target would quickly have the Reserve Bank chasing its own tail, tightening monetary policy (increasing interest rates) in response to an increase in measured inflation which was itself caused by increasing interest rates earlier in the year, and vice versa. Both the negative headline inflation figure we now project and the high figure experienced in the mid-nineties are simply the result of the inclusion in New Zealand's headline inflation figure of interest rates.
Most economists are agreed that interest rates should not be in the inflation figure and many major countries already exclude them. The Government Statistician has agreed to remove them from the official CPI from later this year. Thankfully, this will end the potential for public confusion inherent in the present situation. But in the meantime we ignore the interest-inclusive headline figure, as we have done all decade. Of course, we could prevent the headline inflation figure turning negative very easily - by aggressively increasing interest rates! I don't think anybody is strongly arguing for this!
Growth and inflation
Why not add a growth objective for monetary policy? Surely, the most important objective for all economic policy must be to encourage growth, so why isn't the same true for monetary policy? In a fundamental sense, it is. It's just that we now realise that the most constructive thing that monetary policy can do to encourage growth is to deliver price stability, so that businesses planning what products to produce can distinguish relative price movements from the noise of general inflation, so that savers can sensibly plan how to invest their savings, so that the gross distortions created by the interaction of inflation and a tax system geared to historical cost accounting can be avoided, and all the rest.
We now know that inflation damages economic growth, to say nothing of the social damage it also does through the capricious redistribution of income and wealth which it produces.
It is for this reason that the present Policy Targets Agreement makes it clear that the Reserve Bank is to deliver price stability `so that monetary policy can make its maximum contribution to sustainable economic growth, employment and development opportunities within the New Zealand economy'.
In other words, price stability is not simply an end in itself but a means to an end, and that end is better growth, better employment, and indeed a more just society.
The Governor of the Reserve Bank of Australia, Mr Ian Macfarlane, expressed it well late in 1996 when he said that `the other way of expressing an inflation target is to say that monetary policy is set in a way which lets the economy grow as fast as possible without breaking the inflation objective, but no faster.'1
The Governor of the Bank of England, Mr Eddie George, also put it very clearly in a speech last month:
`We're often described as a lot of "inflation nutters", or even as "pointy-headed industrial hooligans" who don't care at all about real economic activity or jobs... That view... is profoundly and fundamentally wrong.
`The implication, of those who take that view, is that they think there is a trade-off between inflation and the rate of economic growth; they think that if only we would let up a bit on controlling inflation then we could all enjoy higher activity and lower unemployment and rising living standards, which are, of course, the really good things of economic life. At the very least, they say, if you were a bit less manic in your pursuit of low inflation, we could avoid some of the worst damage that is currently being inflicted upon agriculture, upon large parts of manufacturing, and even some service sectors.
`And that might even be true for a time. The trouble is that, in anything other than the short term, it would be likely to mean more rather than less economic damage, and lower rather than higher growth and employment...
`We are every bit as concerned with growth and employment as we are about price stability - as anyone in their right mind must be. Permanently low inflation is a necessary means of achieving growth and employment sustained into the medium and long term.'2
I totally agree.
The control of inflation, and output variability
Most of us now accept that monetary policy can not, by tolerating a little more inflation, buy sustainably faster growth. In other words, keeping inflation low is the best contribution which monetary policy can make to growth in the longer term.
But it is also widely accepted that there may be a trade-off between the variability of inflation and the variability of output. In other words, it may be that attempting to keep inflation at precisely 1.5 per cent at all times might lead the Bank to throw interest and exchange rates around quite vigorously - and output around quite vigorously as a result.
In recent years, the distinction between `strict' inflation targeting and `flexible' inflation targeting has grown up in the economic literature, with `strict' inflation targeters being those who aim to keep inflation very close to a narrow target at all times, and `flexible' targeters being those who are willing to accept a little more variability in inflation outcomes for the sake of reducing the variability of interest and exchange rates, and of economic output.3 I am not sure that any central banks are in fact `strict' inflation targeters, but the distinction is useful in highlighting why all central banks are more or less `flexible' inflation targeters.
The Reserve Bank of New Zealand has never been a `strict' inflation targeter. From the very first Policy Targets Agreement, reasons why actual inflation might legitimately fall outside the target band - the so-called caveats - were recognised. At no stage have we treated the centre of the inflation target band as a fetish.
Early in 1996, after inflation fell outside the then 0 to 2 per cent target by 0.1 per cent in the year to March 1996, I wrote to the Minister of Finance and explained that, although we were confident that inflation would be back within range by the following year, I was not confident that further breaches of the target could be avoided in the immediate future without a very aggressive further tightening of monetary policy. I noted that under all the circumstances I could not recommend such an aggressive further tightening, and indeed noted that tightening to the extent which would have been needed to reduce the risk of further breaches of the top of the inflation target in the immediate future ran the risk of throwing the inflation rate through the bottom of the inflation target further into the future.4
So there is plenty of evidence that we have been `flexible' inflation targeters from the beginning. But I suspect that we may be even more willing to be `flexible' inflation targeters now than we were in the late eighties and early nineties, when we were at the beginning of our inflation targeting experience. This is a direct result of the fact that inflation and inflation expectations are now much more firmly anchored at a low level than was the case a decade ago. At that time, we were coming out of a period of nearly two decades of high and variable inflation, a period during which governments had repeatedly promised to reduce inflation and had repeatedly broken those promises. It was hardly surprising that, initially at least, the public greeted the news of one more approach to delivering low inflation with a high degree of scepticism.
Today, by contrast, even those who criticised the commitment to low inflation initially are persuaded that low inflation is likely to be around for a considerable time to come. Businesses no longer assume that they will be able to increase prices more or less automatically each year. Consumers no longer assume that they must buy this month because next month prices will be higher. Those negotiating wage and salary increases strike agreements consistent with stable prices. Long-term interest rates are nearly back to their levels in the sixties. Even home-buyers are beginning to realise that house prices go down as well as up. As a consequence of these more firmly-anchored inflation expectations, the Bank can afford to respond rather more moderately to the prospect of small changes in the inflation rate, confident that, even if inflation does temporarily depart from the mid-point of the target, this is not likely to produce any damaging longer-term changes in pricing and wage-setting behaviour.
We live in an uncertain world
This anchoring of inflation expectations is of considerable assistance in the way monetary policy operates, and should mean that interest rates may not need to move around as strongly as at times in the past.
But does that mean that the Bank will be able to formulate policy with perfect foresight, and without controversy, in the future? Of course not. Inevitably, sometimes things turn out differently than expected.
For example, 90 day interest rates rose from an average of 4.6 per cent in February 1994 to 9.5 per cent in December that year. At the time, that increase in rates seemed to be ample to contain inflationary pressures. But with the wisdom of hindsight it is probable that we should have tightened policy earlier and more aggressively than we did; had we done so, we might have nipped inflationary pressures in the bud at an earlier stage, and it may have been possible to ease policy earlier than was in fact the case. Not many people complained, in 1993, that we were not tightening policy sufficiently quickly.
Late in 1995, we thought we could see inflationary pressures abating in 1996, and so advised the Government that there seemed no reason from a monetary policy perspective to abandon the tax reductions proposed for mid-1996; in retrospect, we were rather too optimistic, and inflationary pressures continued throughout that year, with the tax cuts and, later, additional government spending adding to those pressures. Again, not many people complained at the time.
As President Harry Truman once said, `Any school-boy's hindsight is better than the President's foresight.'
You would be disappointed if I did not mention the introduction of the Monetary Conditions Index with `bands' at the end of June 1997 as another innovation which, with the benefit of experience, has proved less useful than we had hoped. In the extreme test provided by the Asian crisis, we believe that the MCI with `bands' produced unnecessary volatility in short-term interest rates in the 12 months or so after it was first introduced (that volatility began to diminish in the second half of 1998 as we more explicitly relaxed the `bands' around desired conditions), so in that sense we believe that the MCI was not as helpful in the implementation of policy as we believe the Official Cash Rate will be.
But we have not abandoned the MCI as an indicator of monetary conditions, and we intend to continue using it to illustrate how we see monetary conditions evolving over the period covered by our projections.
Moreover, we do not accept that the MCI `caused' the brief recession in the first half of 1998. The monetary conditions relevant to the level of activity in the first half of 1998 were those in 1996 and early 1997, not those in the second half of 1997. If monetary policy carries any part of the blame for that early 1998 recession, it was monetary policy in 1996 and the first part of 1997 which was to blame; the MCI was introduced at the end of June 1997. In 1996, notwithstanding that inflation was slightly above the then 0 to 2 per cent target and notwithstanding the tax cuts and increased government expenditure, we were projecting inflationary pressures to abate over 1997 and 1998, with a period of positive but below-trend growth. In other words, we were projecting the proverbial soft landing. On this basis, we began easing, in December 1996. What turned a soft landing into something much harsher was the Asian crisis and the drought, both unpredicted and unpredictable in 1996.
In fact, judging from the inflation outcomes it is not at all obvious that monetary policy in 1996 was too tight. CPIX inflation has been close to the middle of the inflation target since the June quarter of 1997, and indeed, had it not been for the widening of the target from 0 to 2 per cent to 0 to 3 per cent at the very end of 1996, inflation outcomes in much of 1998 would have been rather uncomfortably close to the top of the band. On this basis, it is certainly hard to argue that policy in 1996 was too tight.
But there have been mistakes in the past, as I have indicated. And there will be mistakes in the future. Without a crystal ball, this is quite inevitable, given the huge amount of uncertainty which surrounds the way in which policy must of necessity be implemented, namely looking forward one to two years.
There is uncertainty about where the economy is when we begin each projection. The projection which we published last week was finalised on 1 March, and yet the latest comprehensive data we have on economic activity is for a quarter which finished five months earlier, and we know that data for that quarter may be revised, and revised again, for many quarters to come. Of course, we try hard to offset this problem by studying a very large number of more up-to-date statistical series, by talking to scores of businesses all over the country, by talking to audiences in many parts of the country, by assessing the implications of several surveys of business and consumer confidence, and by asking our seven non-executive Directors for their assessment of the economy. As I remarked on a previous occasion, I sometimes suspect that God made economic forecasters to make weather forecasters look good, but at least weather forecasters can look out the window and see what the weather is like today! Economic forecasters do not have that luxury.
There is also a great deal of uncertainty about `how the economy works', in other words about how people and businesses react to circumstances. For example, our latest projections assume that household expenditure will grow more slowly as the economy grows out of the early 1998 recession than it did as we grew out of the recession of 1991/92. We made that assumption because people now carry a great deal more debt, relative to their incomes, than was the case a decade ago. We are assuming that they will take on rather less additional debt over the next few years. That seems an entirely reasonable assumption, but of course nobody can know that with certainty, particularly given that low interest rates of the kind now prevailing have not been experienced in New Zealand for many years.
There is perhaps even more uncertainty about `what lies around the corner'. Will the US share-market continue rising over the next year or two, stabilise at present levels, or fall substantially? The answer to that question, on which we all have views no doubt, will almost certainly have a material bearing on the growth of US consumer spending and potentially therefore on world economic growth. But should we formulate monetary policy on the basis that the US share-market will correct downwards, or will not? Similarly for developments in the Japanese economy. Even within New Zealand, there may be all kinds of totally unpredictable events. Another drought perhaps? Unlikely surely, but certainly not impossible.
All these sources of uncertainty mean that, when we look back on the way in which monetary policy was formulated in March 1999 from the vantage point of, say, 2002, what seemed like well-founded decisions as we finalised our decision on the Official Cash Rate last week may seem hard to justify. Because of the delay between monetary policy action and the effects of that action on the real economy and, through those effects, on inflation, we have no alternative than to formulate policy in full recognition of this inherent uncertainty.
This has several implications, and this is not the place to spell out all of them. But one of the ways in which we try to reduce the risks associated with the inherent uncertainty of monetary policy formulation is by publishing our quarterly projections and the key assumptions on which they are based. (Indeed, we have published considerable information about the model on which our projections are based.) This helps markets to interpret new information as it emerges, and to adjust monetary conditions even before we adjust the Official Cash Rate. Of course this only works because financial market participants understand the objective of monetary policy in New Zealand and, on the basis of our track record over the last decade and more, understand also that we are committed to delivering that objective.
What this means in practice will no doubt be illustrated tomorrow. In our Monetary Policy Statement last week, we indicated that when we finalised our projection on 1 March we expected that GDP would have increased by 0.6 per cent in the December 1998 quarter (seasonally-adjusted). It is possible that, when the figure for December 1998 quarter GDP is released by the Government Statistician tomorrow, it will show growth of exactly 0.6 per cent. Possible, but unlikely. But in an important sense that doesn't really matter. Not only is there ample time for us to adjust policy long before any mis-assessment has the slightest relevance to inflation in the second half of next year, but financial markets themselves may adjust conditions within minutes of the announcement. Thus if the December quarter GDP figure turns out to be appreciably stronger than the Bank projected, and if this is seen by markets as being relevant to the inflation outlook, monetary conditions might well tighten quickly. And vice versa. This is precisely what happened, for example, on the announcement of the GDP figures for the March quarter of 1998 at the end of June, for the June quarter at the end of September, and for the September quarter near the end of December. In each case, monetary conditions moved in a broadly appropriate direction.
At the beginning of my address I asked what New Zealanders can expect of monetary policy in the long term. I would hope that they can expect monetary policy to become so totally predictable as to be boring - stable average prices year after year after year - because that is the best environment for growth in output and employment.
Interest rates will still move up and down to some degree of course, reflecting changes in the balance between savers and borrowers. But the week-to-week volatility of short-term interest rates should be diminished by our move to implement monetary policy through an Official Cash Rate, and our progressively longer track record of low inflation should make it unnecessary ever to push interest rates to the levels seen in the early stages of disinflation in the mid-eighties.
The New Zealand dollar will also fluctuate of course, certainly against individual currencies and even against the trade-weighted index. As I discussed in a speech to the Canterbury Employers' Chamber of Commerce at the end of January, that appears to be a fact of life which nobody has yet found a way of avoiding.
So I suppose that, with interest rates and the exchange rate still moving up and down, monetary policy will never seem totally boring. But with luck monetary policy will at least come to be seen for what it is when it is formulated with skill, namely a useful contributor to a prosperous society, but not nearly as important as a whole host of other matters - policies on education, policies on the labour market, policies on taxation, our access to international markets, the quality of our managers, and all the rest. Those are the things which will really determine our longer-term prosperity.
1 Speech by Mr I J Macfarlane, Governor of the Reserve Bank of Australia, to the CEDA annual general meeting dinner, Melbourne, 28 November 1996.
2 Speech by Mr E A J George, Governor of the Bank of England, at Hertfordshire University,
18 February 1999.
3 See Lars Svensson, `Inflation targeting in an open economy: strict or flexible inflation targeting?', Reserve Bank of New Zealand Discussion Paper, G97/8.
4 Letter to the Minister of Finance dated 19 April 1996, published in Monetary Policy Statement,